Kinetics of dithionite ion utilization and ATP hydrolysis for reactions

Judith F. Rubinson , James L. Corbin , and Barbara K. Burgess. Biochemistry ... Faith E.H. Katz , Xinying Shi , Cedric P. Owens , Simpson Joseph , F. ...
0 downloads 0 Views 873KB Size
WATT

of mammalian cells. The active factor appears to be a protein that interferes with the poly(A)-protein interaction in the complex. We provide evidence that this factor renders the poly(A) susceptible to snake venom exonuclease. The hydrolytic agent in the cytoplasm may be an exonuclease as well, but further work with more purified preparations will be required to clarify its mode of action. The inhibitory effect of exogenous RNA on poly(A) hydrolysis was observed with the cytoplasmic preparation as well as with the snake venom enzyme. Our results indicate the possibility of a control mechanism based on differential susceptibility of the poly(A) segment to degradation in the cytoplasm. Some suggestion of differential poly(A) stability is provided by the data on the time course of hydrolysis in Figure 2. More precise data will be required, preferably with unique m R N A species, in order to verify this possibility. Acknowledgments W e thank Mr. Albert0 Lacardi for excellent technical assistance. References Blobel, G. (1973), Proc. Natl. Acad. Sci. U.S.A. 70, 924. Brawerman, G. (1973a), Mol. Biol. Rep. 1. 7. Brawerman, G. (1973b). International Symposium on Normal

AWL1

BlJRNS

and Pathological Protein Synthesis in Higher Organisms, INSERM, Paris. Brawerman, G., and Diez, J. ( 1 975), Cell 5, 27 1. Brawerman, G., Mendecki, J., and Lee, S. Y. ( 1 972), Biochemistry 11, 637. Greenberg, J. R., and Perry, R. P. ( 1 972), Biochim. Biophys. Acta 287, 36 1. Jeffery, W. R., and Brawerman, G. ( 1 974), Biochemistry 13, 4663. Kwan, S. W., and Brawerman, G. (l972), Proc. Natl. Acad. Sci. U.S.A. 69, 3247. Lee, S . Y . , Krsmanovic, V., and Brawerman, G . (1971a), Biochemistry 10, 895. Lee, S. Y., Mendecki, J., and Brawerman, G. (1971 b), Proc. Natl. Acad. Sci. U.S.A. 68, 133 1. Mendecki, J., Lee, S. Y . , and Brawerman, G. (1972), Biochemistry 11, 792. Penman, S., Vesco, C., and Penman, M. (1968), J . Mol. B i d . 34, 49. Perry, R. P., La Torre, J., Kelley, D. E., and Greenberg, J . R. (1972), Biochim. Biophys. Acta 262, 220. Sheiness, D., and Darnell, J. E. (1973), Nature (London),New Biol. 241, 265. Soreq, H., Nude], U., Salomon, R., Revel, M., and Littauer, U. Z. (1974), J . Mol. Biol. 88, 233.

Kinetics of Dithionite Ion Utilization and ATP Hydrolysis for Reactions Catalyzed by the Nitrogenase Complex from Azotobacter vinelandii? Gerald D. Watt* and Andrea Burns

ABSTRACT: The kinetics of S2042-utilization and A T P hydrolysis during the nitrogenase-catalyzed H2 evolution and acetylene and nitrogen-reducing reactions were studied using a polarographic technique to m0nitor,S204~-concentration. Rate constants for both s 2 0 4 * - utilization and ATP hydrolysis were determined as a function of temperature and corresponding activation energies determined. The activation energy for ATP hydrolysis differs from that for product formation or S2Od2- utilization by 5 kcal/mol above 20 O C and by 25 kcal/mol below 20 OC. The rate law for S 2 0 d 2 - utilization was determined and describes the enzyme catalyzed rate over a

1000-fold variation in s204*-concentration and at least a 100-fold change in ATP concentration. The rate law for S 2 0 4 * - uilization under Nz-reducing conditions a t 25 "C is given by -d([S204*-])/dt = (2.3 X 10-3E~[S2042-]'':[ATP]*)/([ATP]* K I [ATP] K z ) , where ET is total enzyme concentration in mg/ml and K I and K z are equilibrium constants for ATP binding to nitrogenase. The half-order dependence of the rate on S ~ 0 4 concentration ~is interpreted in terms of the equilibrium S 2 0 4 2 - = 2SOz-, in which SO*is the actual electron donor to nitrogenase. A partial mechanism incorporating these results is presented.

N i t r o g e n a s e is capable of reducing a wide range of substrates possessing unique and diverse properties. These include neutral gaseous molecules (N2, N20, and acetylene), relatively large organic molecules (isonitriles, monosubstituted acetylenes, and nitriles), and anions (CN-, N3-).In the absence of any of the above, H 2 0 ( H + ) serves as a substrate. In addition to these reducible substrates, nitrogenase also requires: (1) a source of low potential electrons, with E" more negative than -400 mV (Watt and Bulen, 1974; Evans and Albrecht, 1974);

and (2) MgATP2- which is hydrolyzed to MgADP2- and P,.l Zumft and Mortenson (1975) recently reviewed the properties of nitrogenases from bacterial sources and discussed the meager kinetic results so far reported and the problems associated with studying this aspect of nitrogenase catalysis. The rate dependence on ATP for nitrogenase-catalyzed reactions

Contribution No. 574 from The Charles F. Kettering Research Laboratory, Yellow Springs, Ohio 45387. Rereioed February 5, 1976.

264

BIOCHEMISTRY, VOL.

16,

NO.

2, 1971

+

'

+

Abbreviations used: P,, inorganic phosphate produced from A T P hydrolysis; ATP, adenosine 5'-triphosphate; Tes, N-tris(hydroxymethy1)methyl-2-aminoethanesulfonic acid; Tris, tris(hydroxymethy1)aminomethane; EPR, electron paramagnetic resonance.

DlTHlONlTE

UTILIZATION

A N D ATP HYDROLYSIS BY NITROGENASE

is unclear, particularly a t low A T P concentrations where both sigmoidal (Moustafa and Mortenson, 1967; Silverstein and Bulen, 1970; Bergerson and Turner, 1973) and nonsigmoidal (Thorneley and Willison, 1974; Ljones, 1973) kinetics have been reported. Even less kinetic information is available concerning the rate a t which S ~ 0 4 or ~ -other electron donors are used by nitrogenase. The rate at which reduced products a r e formed by nitrogenase has been reported (Burns, 1969; Silverstein and Bulen, 1970; Shilov and Likhtenshtein, 1971; Ljones, 1973) to be independent of the nature of the substrate for N2 reduction or H2 evolution. Ljones (1973) reported that C N - and N3- inhibited the rate of utilization of S 2 0 4 2 - but attributed this inhibition to protein destruction by these substrates. These results and others (Zumft and Mortenson, 1975) suggest that substrates are reduced a t a constant rate by nitrogenase. The consequence of this is that substrate reduction is not likely to be involved in the rate-limiting step of nitrogenase catalysis. The answer to the question, “what is the rate-limiting step in nitrogenase catalysis”, is thus not clear and becomes important in attempting to understand the mechanism of nitrogenase action. The detailed form of the rate law describing the utilization (production) of the various nitrogenase substrates (products) is important in any discussion of the rate-limiting step of enzyme catalysis as well as in attempts to explain the mechanism of enzyme action. We have previously reported (Watt et al., 1975) that polarographic and calorimetric techniques provide versatile and sensitive methods for detecting nitrogenase activity. These methods have now been developed (Watt, 1976) for the study of the kinetics of nitrogenase activity over a wide range of substrate concentrations and for the determination of the rate laws governing their utilization. This paper describes the results of these kinetic investigations. Experimental Section Materials. Adenosine 5’-triphosphate (ATP) and (hydroxymethyl)methyl-2-aminoethanesulfonicacid (Tes) were purchased from the Sigma Chemical Co. Creatine kinase was purchased from Nutritional Biochemical Corp. Creatine phosphate was synthesized in this laboratory by Dr. James Corbin. Sodium dithionite, Na2S204 (98%), was purchased as a special order from Associated Chemical Co. Ltd., Harrogate, Yorkshire, England. The sealed metal container was opened and subsequently stored in a Vacuum Atmospheres Corp. drybox under oxygen-free argon. Smaller samples of Na2S204 for laboratory use were removed from the stock supply as needed and stored in argon-filled desiccators over anhydrous CaCI2. When stored with these precautions, the Na2S204 remained >95% pure for periods greater than 2 years as measured by amperometric or calorimetric titrations against K3Fe(CN)6 or riboflavin 5’-phosphate. Argon and prepurified nitrogen were passed through a hot quartz tube (600 “ C ) containing copper turnings to remove oxygen prior to use. Preparations of nitrogenase complex from A . vinelandii were obtained as previously described (Bulen and LeComte, 1966, 1972, unpublished results) and possessed activities of 200-300 nmol of Hz evolved per min per mg of complex. ATP/2evalues of 4-5 at 25 O C were usually obtained with these enzyme preparations. Methods. All solutions were prepared anaerobically using airless glassware and transferred with Hamilton gas-tight syringes. N2 reduction was measured as described previously (Bulen and LeComte, 1966). H2 evolution was measured either manometrically or with a Hewlett-Packard Model 5750 re-

search gas chromatograph (molecular sieve 5A, argon carrier gas). Ethylene formation was determined by gas chromatography (alumina or Poropak-N). A creatine phosphate-creatine kinase A T P generating system was used to maintain a constant ATP concentration and to prevent A D P from reaching inhibitory levels. No variation in the rate of S ~ 0 4 utilization ~a t any of the temperatures studied was observed upon increasing the creatine kinase levels by a factor of 4. The P, released was measured as previously described (Bulen and LeComte, 1966) using controls minus Na2S204. The rates of S ~ 0 4 utilization ~by the nitrogenase catalyzed Hz-evolution, acetylene-reduction, and N2-reduction reactions were measured polarographically. The polarographic apparatus and procedure for measuring S z 0 d 2 - concentration have been described (Watt et al., 1975). A standard assay solution of 5.0 ml was used which consisted of Tes or Tris buffer (250 pmol), pH 8, creatine phosphate (260 pmol), creatine kinase (5 nmol), and variable amounts of both S ~ 0 4 and ~ - ATP. The MgZf concentration always exceeded 3.0 m M and had no effect on the rate of S ~ 0 4 utilization ~in the range of 1.0 to 5.0 m M a t a given A T P concentration. Under the above experimental conditions, the MgATP2- complex was the only important ATP-containing species present during the reactions (see Thorneley and Willison, 1974). The rate of S20d2-oxidation in this standard assay solution was measured: (1) at various ATP concentrations maintained constant by the ATP-generating system; (2) at various temperatures; and (3) under various atmospheres (N2, H2, Ar, CO, or acetylene) by adding 0.2 ml of nitrogenase enzyme (1 5-30 mg/ml) to the polarographic cell which contained 5.0 ml of the equilibrated, oxygen-free assay solution. The kinetics of Sz042-utilization was followed by continuously recording the anodic diffusion current of S 2 0 4 2 - until all S 2 0 ~ was ~ - consumed (1 5-25 min at temperatures 2 2 5 “C and 25-45 min at temperatures below 25 “C). Some experiments using a Durrum stopped-flow spectrophotometer followed the oxidation of S2042-optically a t 320 nm. The rate of H2 evolution with limiting Sz042-was measured manometrically (Bulen and LeComte, 1966). The concentration of S ~ 0 4 ~ in -the standard assay solution was determined polarographically and 5.0 ml of this solution was transferred by gas-tight syringe to the manometers, previously degassed and containing enzyme. When equilibrated, the enzyme and assay solution were mixed and readings were taken every minute as the enzyme-catalyzed reaction consumed S 2 0 4 * - and produced Hz. Results Figure 1 shows a polarographic trace of S2042- utilization during the hydrogen-evolution reaction catalyzed by nitrogenase. The anodic current before enzyme is added (arrow) gives the initial S2042- concentration in the assay mixture. The constancy of this current with time demonstrates that no S204*--consuming reactions are occurring in the absence of enzyme catalysis. After enzyme is added an induction period was noted (not shown in Figure 1) in many of the reactions during which no immediate S z 0 d 2 - utilization occurred even though all of the reactants were present for the enzyme-catalyzed reactions to proceed. The time interval of this induction period is very temperature sensitive, becoming shorter a t temperatures above 20 “C and longer below this temperature. The presence of this induction period (with those enzyme preparations which possessed it) depends on the sequence in which the enzyme and reagents of the assay solution are mixed. BIOCHEMISTRY, VOL.

16,

NO.

2, 1917

265

WATT A N D

O

2.0

r

-

c

r I

0

2

4 6 MINUTES

BLRNS

8

IO

0.

12

I I ( i t KI. I : A polarographic trace of the nitrogenase-catalyzed utilization of S?Oj?- during the Hl-evolution reaction. The ordinate is the anodic diffusion current of S20j2- and is directly proportional to S2Oj'- concentration ([S:OJ?-] = 1.375 X 10-Jia204?-). The reaction is initiated by adding enzyme to the cell (arrow) containing all reactants necessar) for the nitrogenase-catalyzed, HZ-evolution reaction (see text for details). The insert is a plot of [SzOj'-]'!' against time i n minutes.

L U 3.20

3.30

3.50

3.40

+x

103

3: Arrhenius plots for the one-half order rate constants ( k , , z ) for S204?- utilization ( 0 )and the rate constants ( k , , , ~ p ) for A T P hydrolysis (0). T h e A T P concentration is 5 X M a t all temperatures and is maintained constant by the ATP-generating system. E , for A T P hydrolysis ( 0 )was found to be 17.2 kcal/mol, while values of 12.3 and 35 kcal/mol were found for the biphasic SzO4'- utilization curve ( 0 ) . IICitKt

i

I

J 24

Time

(rnin)

I I ( ; [ K I 2 : The nitrogenase catalyzed H?-evolution reaction a t 25 'C under the conditions of limiting [S20a2-]. The upper set of curves (0,0 ) u i j s obtnined by adding 0.20 ml of nitrogenase complex preincubated ( I O niin) w i t h MgATP?- to manometers containing the ATP-generating s!stem and [S201'-] initially at 5.0 X IO-3 M. The lower set of curves ( 0 ,a ) *;IS obtained a5 above. except preincubation of the nitrogenase cotiiplcx was omitted. The ratio of the total H' measured to total S ~ 0 a ' 0xidi7cd in the four kinetic experiments shown varied from 0.96 to I .ox.

those of Figure 1 were obtained when the enzyme-catalyzed utilization of SzO4'- in identical reaction mixtures was run under N2, Ar, H2, CO, or acetylene atmospheres. Figure 2 shows that similar kinetic behavior is seen by measuring the rate at which Hz is evolved by the enzymecatalyzed, H2-evolution reaction. Preincubation of enzyme and MgATP*- causes H2 evolution to begin immediately, whereas an induction period is seen in the absence of preincubation. Once H2 evolution begins, its rate occurs uniformly at first but then decreases with time as S20~'- is consumed. The rate of H ? evolution for the last part of the curve in Figure 2 appears to be a function of [S204'-]'~2, i.e., d[Hz]/dt = k [ S ~ 0 4 ~ I/', -] whereas the rate of Hl evolution for the initial part of the curve is independent of [S2042-] (zero order in [Sz04*-]). It is clear from Figure 2 that 1 mol of Hz is evolved per mol of S 2 0 d 2 consumed. Variation of Rate with Temperature. The one-half order rate constants at various temperatures for sz04'- utilization by the nitrogenase-catalyzed, H2-evolution reaction, maintained at 5 m M ATP by the ATP generating system, are given in Figure 3 in the form of an Arrhenius plot. The rate constants for phosphate (P,) production resulting from ATP hydrolysis at the same temperatures are also included in this figure. The rate constants for ATP hydrolysis a t the various temperatures studied were calculated from eq 1

I t occurs if the enzyme is added last to the assay mixture, as was done for most of the results reported here. I f enzyme and MgATP'- are preincubated before addition to the assay solution in the polarographic cell, the induction period is eliminated and SzO4'- utilization begins immediately. Preincubation of enzyme with other constituents of the assay solution were ineffective in eliminating the induction period. Although adding enzyme last to the assay mixture introduces some uncertainty in analyzing the first minute or so of which relates the rates of S704'- utilization, measured in this the enzyme-catalyzed reaction, the remaining 15 to 30 min are study, to the ATP/S2O4:- value determined previously as a more than adequate to properly evaluate the rate of reaction, function of temperature (Watt et al., 1975). The one-half order and the overall convenience and reproducibility gained by this rate constants for S 2 0 4 2 - utilization produce a biphasic plot, procedure makes the omission of this initial part of the curve while the rate constants for P, formation form a single straight justifiable. line. Following the induction period (when present) S ~ 0 4 ~ - Activation energies of +35 and 12.3 kcal/mol for S ~ 0 4 ~ utiliiation begins and the anodic current ( S ~ 0 4 concentra~utilization are obtained from the 10-20 O C and the 20-35 "C tion) decreases with time as shown in Figure 1. Treatment by temperature regions, respectively. These results a r e in agreeusuai kinetic procedures indicates that this curve corresponds ment with those previously reported for Azotobacter vinelandii to a one-half order reaction with respect to S 2 0 4 2 - concennitrogenase (Burns, 1969; Shilov and Likhtenshtein, 197 1) tration. This dependence is strictly obeyed to the conclusion obtained from the rates of product formation. of the reaction (see insert to Figure 1 ) . Curves identical with The Arrhenius plot for the rate of P, production ( A T P hy-

+

DlTHlONlTE UTILIZATION

14.0

A N D ATP H Y D R O L Y S I S BY N I T R O G E N A S E

r !

0-

Y)

P X

+ u

B

0

0.2

0.4 [ATP]

0.6

0.8

I .o

2.0

mMolar

FIGURE 4: The rate of the nitrogenase-catalyzed, H2-evolution reaction at 25 ( 0 )and 30 'C (0)as a function of A T P concentration in the pres~ence of the ATP-generating system and 10-20 m M initial S ~ 0 4 concentrations. The ordinate is the rate (molar Sz04*-/min) at which S 2 0 4 2 is consumed by the nitrogenase-catalyzed reactions as measured polarographically.

drolysis) catalyzed by nitrogenase yields an activation energy of 17.2 kcal/mol. This result taken together with those for S 2 0 d 2 - utilization discussed above clearly indicate that the rates of ATP hydrolysis and electron transfer from S ~ 0 4 to ~products catalyzed by nitrogenase are affected differently by variations in temperature. In fact, at 10 O C the rate of S2042utilization is barely detectable, while ATP hydrolysis occurs at an easily measurable rate. At this temperature or below, nitrogenase is essentially an ATPase incapable of transferring electrons to substrate. The rate of s~o4~utilization at 25 O C for acetylene and nitrogen reduction produces curves identical with that for H2 evolution shown in Figure 1. Identical rate constants with the same temperature variations (same activation parameters as those for H2 evolution) were observed from acetylene and nitrogen reduction. Rate Law f o r S ~ 0 4 and ~ - ATP Utilization during Hz Evolution. Figure 4 shows the relationship between the rate of S2042- utilization and ATP concentration for the H2-evolution reaction. Sigmoidal curves are clearly evident a t both 25 and 30 'C. Above 0.3 m M the rate becomes insensitive to increasing [ATP], indicating enzyme saturation with respect to ATP. Below 0.01 mM ATP, no S 2 0 4 2 - utilization could be detected, even though all components necessary for nitrogenase activity were present. These same results were obtained under N2-reducing conditions (1 atm N2), where a more complete and thorough study was done (see later). The data in Figure 4 at initial S ~ 0 4 concentrations ~greater than 10 mM are fit empirically by eq 2

+

-d[S2042-] - ETA [S2042-]1/2[ATP]2 (2) dt [ATP]' Ki [ATP] K2 where K i , K2. and A at 25 O C are numerically equal to 5.05 X M-I, 2.8 X M-I, and 2.35 X ml min-l mg-I, respectively. ET is the concentration of nitrogenase complex expressed in mg/ml. An equation of this same form with A = 3.1 X ml min-l mg-', K i = 2.0 X 10-5 M-I, and K2 = 1.3 X M-I describes the rate of S2Od2utilization at 30 "C. Rate Law f o r S2Od2- and ATP Utilization during N2 Reduction. The rate of S Z O ~ utilization ~by nitrogenase as a function of S 2 0 4 2 - concentration under Nz-reducing conditions (1 atm N2) and at various A T P concentrations is shown

+

+

[S,Oil

rnMolar

FIGURE 5: The rate of s 2 0 4 ' - utilization (molar S204*-/min) at 2 5 O C under Nz-reducing conditions (1 atm N2) as a function of both S2Od2and ATP concentrations. The solid line is calculated from eq 2 and the data points are obtained experimentally. The insert is a double-reciprocal plot of l/rate vs. I/[S204z-]'/2. The A T P concentrations beginning with the lower curveand increasing are: 5 X 7.5 X 1.3 X IO-4, 5 X and I X IO-' M.

in Figure 5. The solid lines are calculated from eq 2 and provide a good fit to the experimental points over a 100-fold variation in ATP and a 1000-fold variation in s ~ 0 4concentration. ~The close agreement between experimental and calculated results provides a basis for using eq 2 for computer-assisted analysis of our kinetic data. Although eq 2 was derived under H2 evolving conditions, it also accurately describes the rate of S204*- utilization for N2 and acetylene reduction and indicates that the rate of enzyme catalysis is independent of the type or concentration of the reducible substrate. Therefore, the reducible substrate must either interact with the enzyme prior to or independently of the ATP and S 2 0 4 2 - reactions. Once the enzyme is poised by the ATP-S2042- reactions, it is irreversibly committed to react and any reducible substrate which is available undergoes reduction. The curves in Figure 5 are truly parabolic as is seen from the insert where the reciprocal rate is plotted against 1 / [S2042-]'/2. The half-order dependence of the rate on [ S ~ 0 4 ~ -suggests ] that S02- is the actual reducing agent, formed by eq 3 (Lambeth and Plamer, 1973; Creutz and Sutin, 1974). ki

s~04~k-i

2s02-

K = 1.4 x 10-9

(3)

M-1,

kl = 1.7 S - I , k-i = 1.2 x 109 M-L s - l

The slopes of the lines in the insert are clearly functions of the ATP concentration. This function was evaluated and corresponds to eq 4. slope = 425

[ATPI2

+ K I [ATP] + K2

(4) [ATP] A replot of these slopes against the reciprocal of the ATP concentration gives a straight line from whose intercept a value of 2.35 X M i l 2 ml rnin-' mg-i was obtained. This Kso2-/ VsoZ-value is well determined but individual values of Kso2- and Vso2-could not be evaluated from our data because of the indeterminate condition (both &02- and l ' s ~ ~ BIOCHEMISTRY, VOL. 16, NO. 2 , 1977

267

,./.,

I BO

I601

(b)

120t

80 60

I00

200

300

;"

40C

0 4

8

12

I6

20

IO-^ I / A T P X 10" I 1 0 1 K I , 6: (a) A plot of I /rate a&ainht I /[ATP] for nitrogenase catalysis i ~ t25 "C. pH 8.0 and I a t m h?.The solid lines are calculated from eq 2. The points ( 0 )are representative experimental data. (b) Same as a. except I /rate i s plotted against I /[ATP]'. The insert is a replot of the intercepts in cither B or b plotted against I /[S:032-] '?. The S2012- concentrations beginning with the l w e r curve and increasing are: 5 X IO-'. 7.5 X I X 1W4. 5 X I X IO-'. and 5 X IO-' M. I / A T P ~x

'

appear to be infinite) resulting from all lines in the insert to Figure 5 intersecting at zero. The intersection of all lines at zero indicates that nitrogenase never becomes saturated with S 2 0 4 ' - (or SOz-) even a t infinite SrO4'- concentrations. In other words, Sol- reacts "chemically" (without enzymereductant complex formation) with nitrogenase. Plots of reciprocal rate vs. [ / [ A T P ] and ]/[ATP]' are shown in Figures 6a and 6b, respectively. A replot of the intercepts of these figures vs. 1 / [ S 2 0 4 ' - ] l l 2 is shown as an insert i n Figure 6a. A straight line is obtained for both plots which intersects at the origin. The Kso:-/ Vs0:- value is obtained from the slope of these plots, but separate values of Vs0,- and Ksol- again appear to be infinite. Discussion Nitrogenase requires three substrates for activity: ( 1 ) a reducible small molecule (Nz, CN-, acetylene, H 2 0 , etc.), (2) a source of low potential electrons, which in the present study is supplied by S~04'-, and (3) MgATP2- which becomes hydrolyzed to MgADP2- and P,. The rate laws describing the utilization of each of these substrates and the formation of products are necessary information in eventually understanding the mechanism by which nitrogenase functions. Certain aspects of each of these substrates interacting with nitrogenase are discussed separately and then later discusse,d collectively i n terms of the mechanism of nitrogenase action which they suggest. Small Molecule Substrates. Previous kinetic studies of the N2-reducing and H2-evolving reactions catalyzed by Azotobacter iinelandii nitrogenase (Burns, 1969; Silverstein and Bulen, 1970) have shown that the rate of production of reducing equivalents (measured by the rate of product formation) are identical. Ljones (1973) reported that the rate of S 2 0 4 ' - utilization catalyzed by Clostridium pasteurianum nitrogenase was constant whether under N2-reducing or H2-evolving conditions. We have confirmed the above results by measuring the rate of S2Od2- utilization polarographically (for the H2-evolution reaction, the rate of product formation was also measured) and have shown further that acetylene reduction also occurs at this same rate. These three substrates [N2, H 2 0 ( H + ) , acetylene] are sufficiently different in their properties, reduction products, and requirements for electrons that identical rates of reduction would not be expected. Consequently, the rate of substrate reduction is not rate limiting

268

BIOCHE~.IISTRY.VOL

16, NO

2, 1971

as has been suggested by Zumft et al. (1974) but instead this distinction belongs to some other reaction which controls the rate. S104'Substrate. The kinetics of enzyme catalysis was measured by determining the rate of S104'- utilization in the presence of a constant A T P concentration maintained by the ATP-generating system. Under these conditions the dependence of the rate on ATP concentration should be zero order i f the ATP concentration is maintained constant and the observed order of the reaction is that due to S20jZ-.The A T P concentration remained constant throughout the reaction as evidenced by: (1 ) the strictly obeyed one-half-order reaction for the entire reaction interval (Figure 1) and (2) the lack of variation in rate as components of the ATP-generating system were increased by a factor of 4. This one-half-order dependence on [Sr04?-]is important in understanding how low-potential electrons are used by nitrogenase. Lambeth and Palmer ( I 973) and Creutz and Sutin ( 1974) have discussed the different mechanisms by which S104'- acts as a reducing agent. Lambeth and Palmer (1973) concluded from a kinetic study of the reduction of a series of redox proteins by S204'- that the almost exclusively preferred mechanism involves SO:- radicals with rate constants in the range 1 O7 M-' s-l. The dependence of the rate on S2Oj'- for these redox reactions was reported to be one-half order. The recent kinetic study reported by Thorneley et al. (1976) indicates that SO?- reacts with dye-oxidized Fe protein from Klebsiella pneumoniae with a rate constant > I O 8 M-' s-' The observed one-half order dependence of [ S ~ 0 4 ~ utiliza-] tion reported here for nitrogenase catalysis is therefore consistent with SO?- being the reductant for the Fe protein. Consistent with this view, we have found that the addition of Fe protein to reaction mixtures being catalyzed by the Azotobacter nitrogenase complex increases the rate of SzO4'utilization, although at present the exact dependence of the rate on the Fe protein concentration has not been determined. Whatever the electron accepter is on nitrogenase, we consider its acceptance of an electron from SO?- to be essentially irreversible for two reasons. The first is that the difference between the midpoint potential of S204'- and nitrogenase at pH 7 is on the order of 200 mV for an n = 2 process in favor of reduction. The second and probably the most important is that the product of SrOj2- oxidation is SO?'- which at neutral or alkaline pH is not easily reduced back to SlO4'- (Meites, 1955). The rate of Hz evolution in Figure 2 a t low S.04'- concentration becomes curved and can be shown to depend on the square root of the S204'- concentration. However, at high S70~'- concentration the linear relationship in Figure 2 indicates that the rate of Hz evolution is independent of S ~ 0 4 ' concentration. These results are similar to those of Ljones and Burris ( 1 972) who reported that the rates of S 2 0 4 ' - utilization and H 2 evolution using Clostridium pasteurianum nitrogenase were independent of S?O4?-concentration. Only a t very low S 2 0 4 ' - concentrations is loss of linearity observed in their data for H1 evolution which seems to correspond to a square root dependence of the rate on SlO4'- concentration. The result reported here and those reported by Ljones and Burris (1972), indicating that the rate of product formation is independent of S104z- concentration, suggests that the rate of product formation may be limited by the rate at which ATP and nitrogenase interact and not by S.04'- interaction except when the latter becomes limiting. MgA TP'- Substrate. According to Thorneley and Willison ( 1974). MgATP'- is the form of A T P used by nitrogenase I

.

D l T H I O N l T E U T I L I Z A T I O N A N D ATP H Y D R O L Y S I S BY N I T R O G E N A S E

while Mg2ATP is inactive. Our experiments were designed such that MgATP2- was the predominant ATP-containing species thereby minimizing complicating effects on the observed rates. Our results show definitely that sigmoidal kinetics occur for the acetylene-reducing, Nf-reducing, and H2evolving reactions catalyzed by Azotobacter vinelandii nitrogenase. Silverstein and Bulen (1 970) and MDustafa and Mortenson (1967) had previously reported this effect for nitrogenase from Azotobacter vinelandii and Clostridium pasteuranum from rates of H2-evolution and N2-reduction reactions but the results of the present study using more sensitive methods unequivocally demonstrate this behavior. Studies using nitrogenases from Clostridium pasteuranum (Ljones, 1973) and Klebsiella pneumoniae (Thorneley and Willison, 1974) have indicated that nonsigmoidal kinetics were observed. These studies were not performed a t A T P concentrations as low as those in this study nor with the sensitivity to detect reaction as was available to us. W e suggest that these previous studies could have overlooked this effect using these other enzymes. For example, the Fe protein from Clostridium pasteuranum is known to bind two A T P molecules with dissociation constants of 17 pm (Tso and Burris, 1973). Using these data, the concentrations of free Fe protein and its ATP-containing forms can be calculated as a function of ATP concentration. These calculations suggest that sigmoidal behavior should be observable a t 0.02 m M A T P or below, far lower in ATP concentration than the lowest ATP concentration (0.05 m M ) used by Ljones (1973). W e have not determined why some of our enzyme preparations give an induction period before using S 2 0 b 2 - and why some do not. The induction period can be eliminated by preincubation with MgATP2- but the change which A T P accelerates is not known. The temperature dependence of this induction period parallels qualitatively the activation parameters given in Figure 3 and suggests the two effects are related. A possibility is that a definite sequence of substrate addition (an ordered sequence) to the nitrogenase enzyme must occur. For example, if A T P must add first to the enzyme followed by reduction by S2Ob2-, then an induction period might be expected to occur as the enzyme and A T P interact. Activation Energies. Burns (1969) first reported that nitrogenase from Azotobacter vinelandii gave a biphasic Arrhenius plot with a point of intersection a t 21 O C . Activation energies of 14.6 and 39 kcal/mol were found above and below this temperature. Similar results were reported by Shilov and Likhtenshtein (1971). Silverstein and Bulen (1970) studied the kinetics of nitrogenase from Azotobacter vinelandii as a function of A T P and from their data and the variation of ATP/2e- reported by Hadfield and Bulen (1 969) concluded that two pathways for A T P hydrolysis existed with activation energies differing by 4.6 kcal/mol. One pathway leads to A T P hydrolysis only while the other produces A T P hydrolysis with concomitant electron transfer. Figure 3 confirms the results of Burns (1969) by showing two temperature regions for nitrogenase-catalyzed electron transfer activity with activation energies of 12.3 and 35 kcal/mol. Figure 3 shows further that only a single slope occurs for ATP hydrolysis with an activation energy of 17 kcal/mol. A single straight line could be drawn through the A T P hydrolysis data given by Burns (1 969), instead of the two lines he drew, which gives a reasonable fit of his data. This line yields an activation energy of 20 kcal/mol, in fair agreement with our result of 17 kcal/mol from Figure 3. More recently, Thorneley et al. (1975) have determined the temperature dependence of the rate of acetylene reduction by a 1:1 complex of Klebsiella pneumoniae Fe protein and FeMo

protein and find no break in the resulting Arrhenius plot. An activation energy of +19 kcal/mol results. Our results differ from Thorneley et al. in clearly indicating a sharp break in the Arrhenius plot for the reduction of N2 and acetylene reduction and for H2 evolution. The two processes postulated by Silverstein and Bulen (1970) differing in activation energy by 4.6 kcal/mol in the temperature range 20-40 'C are seen from Figure 3 to be ATP hydrolysis (1 7 kcal/mol) and electron transfer (1 2.3 kcal/ mol). Because A T P hydrolysis has a larger activation energy than electron transfer, an increase in temperature produces a correspondingly greater increase in the rate of A T P hydrolysis than in the rate of electron transfer to form products. The rate of increase in ATP hydrolysis with temperature over that for electron transfer causes the ATP/2e- value to increase with temperature as reported by Hadfield and Bulen (1969) and Watt et al. (1975). The molecular basis for these effects is not .known a t present but protein modification with temperature has been suggested (Watt et al., 1975). Below 20 'C, a similar situation occurs except that the activation energy for electron transfer is now greater than that for A T P hydrolysis. In fact a t 10 O C electron transfer to products is barely detectable, while A T P hydrolysis proceeds at an easily measurable rate. At this temperature, nitrogenase approaches the condition of being simply an ATPase. Because electron transfer is so drastically retarded below 20 O C , while ATP hydrolysis continues, the ATP/2e- ratio rapidly increases with decreasing temperature as was found by Watt et al. (1 975). Again, the molecular basis for these observations is not known but conformational changes affecting electron transfer seem plausible. Mechanistic Interpretation. The reciprocal plots in Figures 6a and 6b show that the point of intersection on the ordinate is a function only of the S 2 0 4 * - concentration. At these intercepts the concentration of A T P is infinite and the enzyme is fully saturated with ATP. The rate of the enzyme reaction under these conditions of saturating ATP depends only upon the square root of the S204*- concentration and as has been previously discussed indicates that SO>- is probably the active reductant. The reciprocal plot of Figure 5 shows that variation in the A T P Concentration has no effect upon the intercepts, all of which intersect a t zero, but its variation does alter the slope of the lines in a manner given by eq 4. This behavior indicates that A T P and enzyme are in thermodynamic equilibrium (Cleland, 1963) and that a change in the A T P concentration specifically alters the net rate of the reaction by altering the concentration of the enzyme-ATP complex which then reacts with SOz-. These two kinetic patterns indicate that A T P binding occurs first followed by reduction by SOf- as given by the following ordered sequence of events where E and E' are probably the oxidized and reduced forms of the Fe protein component of nitrogenase, respectively. E

+ 2ATP +E(ATP)2

(5)

k-i

E(ATP)2

+ SO>- -% E'(ATP)2 + S032k3

E'(ATP)2 -products E(ADP)2

+ E(ADP)*

+E + 2ADP

(6) (7)

k4

(8)

ks

Reaction 5 could be written as two successive steps each involving the addition of one A T P because A T P is probably BIOCHEMISTRY, VOL.

16, N O . 2, 1977

269

W A T T

binding to interacting sites on the protein thus making the two steps kinetically inequivalent. The relationship involving two ATP molecules per SOz- given by reactions 5 and 6 follows from the rate law, eq 2, and in addition is supported by: (1) the stoichiometric relationship of ATP/e- z 2 (Watt et al., 1975, and references therein); (2) binding experiments showing two ATP molecules binding to reduced Fe protein (Tso and Burris, 1973); and (3) EPR studies of single electron reduced Fe protein in the presence of ATP (Zumft et al., 1973). We have written reaction 6 as an irreversible reaction in accord with reasons given above. By analogy to reactions of Sol- with redox proteins, some closely related to the proteins of nitrogenase, reaction 6 could have a rate constant as large as I O 7 M-' S - I (Lambeth and Palmer, 1973). Indeed, Thorneley et al. ( 1 976) have found the rate constant for reduction of dye-oxidized Fe protein by SO*- to be > I O 8 M - ' s-l. The one-half-order rate constant of 2.3 X lo-' M I / ? ml min-' mg-' determined in this study for S204?- utilization by nitrogenase can be converted into a value of 3 X I O 5 M-' s - I for the utilization of Sol- by nitrogenase using the equilibrium constant for reaction 3. Though the conditions of the two experiments differ somewhat, it seems reasonable to conclude that the reduction of the Fe protein under turnover conditions is slower by a factor of > I O 3 than reduction of isolated Fe protein. The mechanism depicted by reactions 5 to 8 and the above discussion can be summarized by eq 9 ATP

ATP

h//k-,

fi2J/h-;

SO,-

SO,-

H? k,!

E(ATPX E(ATP), E(ATP),

2ADP k,llk-:

(9)

E(ADP), E

which is of the same form as the so-called Theorell-Chance mechanism (Cleland, 1963) in which no central complex involving both ATP and SO*- exists. The general rate equation for this mechanism has been derived (Cleland, 1963) and should contain terms in the denominator involving both [Sol-] and [SO*-][ATP] in terms of mechanism 9. The rate equation (eq 2) which provides a good fit to our data (see Figure 5 ) contains no such terms. The reason for their absence may be that the combination of rate constants which makes the coefficients to the [SOz-] and [SOz-] [ATP] terms and the very low concentration of Sol- ([Sol-]= (1.4 X [ S ~ 0 4 ~ - ] where 1.4 X is the dissociation constant for S*04l-) makes these terms small in comparison with the other terms in the denominator of the derived rate equation. The and Kz = 2.8 X of eq 2 are constants K I = 5 X identified as the dissociation constants for ATP interacting with nonreduced nitrogenase and are in fair agreement with corresponding values reported (Tso and Burris, 1973) for ATP ( K , = 1.7 X I O p 5 and K 2 = 3 X interacting with S2041- reduced Fe protein from Clostridium pasteuranium. Acknowledgments Appreciation is expressed to Dr. W . W . Cleland a t the Biochemistry Department of the University of WisconsinMadison for many helpful discussions and for the analysis of some of the kinetic results. Appreciation is also expressed to Drs. E. I. Stiefel, W . E. Newton, J. W. McDonald, and J. L. Corbin for helpful comments during the preparation of this manuscript.

270

BIOCHEMISTRY, VOL.

16,

NO.

2, 1917

AND

HLR\S

References Bergersen, F. J., and Turner, G . L (1973), Biochem. J . 131, 61. Bulen, W. A,, and LeComte, J. R. 1966), Proc. Natl. Acad. Sci. U.S.A. 56, 979. Bulen, W. A,, and LeComte, J. R. ( 972), Methods Enzymol. 24B. Burns, R. C. (l969), Biochim. Biophys. Acta 171, 253. Cleland, W . W. (1963), Biochim. Biophys. Acta 67, 104. Creutz, C., and Sutin, N . (1974), Inorg. Chem. 13, 2041. Eady, R. E., and Postgate, J. R. (1974), Nature (London) 249, 805. Evans, M. C . W., and Albrecht, S. L. (1974), Biochem. Biophys. Res. Commum. 61, 1187. Hadfield, K. L., and Bulen, W. A. (1969), Biochemistry 8, 5103. Lambeth, D. O., and Palmer, G. (1973), J . Biol. Chem. 248, 6095. Ljones, T. (1 973), Biochim. Biophys. Acta 321, 103. Ljones, T., and Burris, R. H. (1972), Biochim. Biophys. Acta 275, 93. Mahler, H. R., and Cordes, E. H. (1966), Biochemical Chemistry, New York, N.Y., Harper & Row, p 259. Meites, L. ( 1 9 5 9 , Polarographic Techniques, New York, N.Y., Interscience, p 282. Moustafa, E., and Mortenson, L. E. (1967), Nature (London) 216, 1241. Postgate, J. R. (1971), The Chemistry and Biochemistry of Nitrogen Fixation, London, Plenum Press, p 167. Shilov, A. E., and Likhtenshtein, G. I. (1971), Ser. Biol., Monogr.-Organ. Los Estados Am. Programa Reg. Desarrollo Cient. Tecnol. No. 4, 5 18. Silverstein, R., and Bulen, W. A. (1970), Biochemistry 9, 3809. Smith, B. E., Lowe, D. J., and Bray, R. C. (1973), Biochem. J . 135, 33 1. Thorneley, R. N . F., Eady, R. R., and Yates, G. (1975), Biochim. Biophys. Acta 403, 269. Thorneley, R. N. F., and Willison, K. R. ( 1 974), Biochem. J . 139, 211. Thorneley, R. N . F., Yates, G. M., and Lowe, D. (1976), Biochem. J . 155, 137. Tso, M.-Y. W., and Burris, R. H. (1973), Biochim. Biophys. Acta 309. 263. Watt, G. D. (1976), Proceedings of the IInd International Symposium on Nitrogen Fixation, Sept, 1976, Newton, W. E., Postgate, J. R., and Rodriguez-Barrueco, C., Ed., London, Academic Press. Watt, G. D., and Bulen, W . A. (1974), International Symposium on Nitrogen Fixation, Newton, W. E., and Nyman, C. J., Ed., Pullman, Wash., Washington State University Press. Watt, G. D., Bulen, W. A,, Burns, A., and Hadfield, K . L. (1 975), Biochemistry 14, 4266. Yates, M. G., Thorneley, R . N. F., and Lowe, D. J . (1975). FEBS Lett. 60, 84. Zumft, W. G., and Mortenson, L. E. (1975), Biochem. Biophys. Acta 416, 1. Zumft, W. G., Mortenson, L. E., and Palmer, G. (1974), Eur. J . Biochem. 46, 525. Zumft, W. G., Palmer, G., and Morterson, L. E. (1973), Biochim. Biophys. Acta 292, 413.